.png){kind=icon}
.png){kind=icon}
Return
A Localization Method Based on Multiarea Selection and Error Compensation for Indoor Wireless Sensor Network
.png){kind=icon}
PRE
AI
DOI：10.1109/TIE.2024.3383017.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Indoor localization technology is affected by the combination of random and systematic errors, which results in the ranging error not obeying a single distribution over a large monitoring area, making it difficult to perform comprehensive modeling. This causes the accuracy problem of indoor positioning technology and becomes the most challenging research hotspot in recent years. To address the aforementioned issue, this article proposes a novel method, which converts a large monitoring area into multiple independent areas with relatively stable errors by setting precorrection points. Then, an area selection model is constructed using Optuna optimized light gradient boosting machine (LightGBM), while the ranging error compensation model is created by integrating multiple deep neural networks (DNNs) according to the error data characteristics within each area. Consequently, a new area selection and ranging error compensation (ASEC) network is constructed to compensate the ranging error of the tag node. Finally, the African vulture optimization algorithm (AVOA) is combined to solve for the location coordinates.As verified by practical experiments, the proposed indoor localization method can significantly reduce the ranging error while maintaining excellent robustness, thus obtaining higher localization accuracy with a localization error of no more than 0.034 m.
Keywords:
African vulture optimization algorithm (AVOA)
area selection and ranging error compensation (ASEC)
localization ultrawideband (UWB)
Journal
IF:
7.2
Papers:
1.8W
Citations:
9.8W